Cartography: Module 5 -- Spatial Statistics

     This week was another foray into the ESRI Virtual Campus. I completed the course, “Exploring Spatial Patterns in Your Data Using ArcGIS.” This course focused on the analysis of data using spatial statistics tools found in the Spatial Statistics Toolbox and the Geostatistical Analysis extension. Upon finishing the course, I passed a quiz and was awarded a certificate. What follows is  a brief discussion of the tools used and a map that shows some of what I did within the course. 

     A visual assessment of mapped data is a preliminary analytical step. You can see spatial patterns in your data or note a lack there of. These visual trends are fine but a more in depth evaluation of the data is needed. Spatial statistics allow you to examine the characteristics of your data and leads to a richer analysis than can be provided from a visual perusal. For instance, you may not be able to spot an outlier or if that outlier is affecting your data in any way. With the spatial statistics tools provided in ArcMap you can find the median center, mean center, and directional distribution of the data. These tools are found in the Spatial Statistics Toolbox. To take the analysis further you can use the tools located in the Geostatistical Analyst extension. This toolset will display your histogram, QQ plot, semivariogram, Voronoi map, and the global trends within the data. These tools will help you discover if your data is normally distributed, their frequency and variation, and the presence of outliers. In addition, you can see a 3D trend analysis (that you can rotate in real time...pretty neat) and if your data is spatially autocorrelated. 

     The map I am displaying is a result of the first exercise in the course. The goal of the exercise is to examine the spatial distribution of data. For this exercise, I examined weather monitoring stations in western and central Europe. The mean center, median center, and directional distribution ellipse are all displayed. The mean center, represented by the purple diamond, is the average location of the dataset. From this calculation we can see that while there seem to be some clusters of weather stations, there is enough of a dispersion to put the median roughly in the center of western and central Europe. The median center, represented by the orange cross, is the middle value of all the locations. The median and mean centers are close but indicate that the data impact them differently. The directional distribution runs east to west indicating that more of the stations are distributed in this direction than the are to distributed north to south. 

     As far as design, the map example provided is very busy and full of information. I placed the legend in an area that seemed to be less crowded and would not hide any information. I also tried to place the north arrow and authorship in uncluttered areas. The source information has some overlay issues but it is embedded in the layer. Attempts to alter the citations were not allowed by the permissions set on the layer. I made the distributed information take up as much of the map as possible without losing geographic context. I also designed thematic symbols to be visually weighted. 

This is a map displaying the data examined and analyzed as part of
an ESRI virtual training course in spatial statistics. It shows western and central Europe
and the location of weather monitoring stations throughout the region. The mean center,
median center, and directional distribution ellipse are also displayed.